Concurrent business and distribution strategy planning using Bayesian networks

Business and distribution strategy planning are usually carried out in a sequence. A company first devises a business plan and then a distribution strategy able to accommodate it. The separation in planning can lead to a sub-optimal decision. We propose a method of how to concurrently plan both strategies, using a Bayesian network. We present three modifications of our concurrent optimization model which are based on different optimization objectives - distribution strategy costs minimization, revenue maximization, and profit maximization. The derivation of all model modifications and the collection process of the required inputs are described in detail. The presented model is tested on a business case of the company Pilsner Urquell, a world-renowned brewery based in Pilsen, Czechia. Using the company's historical data from 01/2017 - 12/2017, we design the cost-optimum distribution strategy in the Czech market for the years 2018 - 2020. Our results are then compared with the real company development over the same period. With our model, we show that the company could have selected a more cost-effective distribution strategy in 2017.